Comparison of Restricted BLUP Selection Procedures Supporting Swine Breeding in a Closed Herd

Abstract

Restricted selection procedures aimed at achieving predetermined relative changes for all traits are widely used in swine breeding programs in closed herds in Japan. The present study compared responses to selection made by using different restricted selection procedures in a simulated closed herd. The base population consisted of 10 males and 50 females. Three traits (average daily gain, backfat thickness, and loin eye muscle area) of 10 generations were simulated. Selection was made on the basis of an index of individual phenotype for desired gain (RSI) and restricted best linear unbiased prediction of breeding (RBLUP) values. Two types of restriction for RBLUP were imposed on breeding values:the same restrictions on all animals (AR-BLUP) and restriction on only some animals (PR-BLUP). Ten relative desired changes were used in each restricted selection procedure. The cumulative response to selection was greater for selection on RBLUP than on RSI, as expected. The cumulative genetic gain with PR-BLUP was slightly higher than that with AR-BLUP, but no differences between them were statistically significant at generation 10 in any relative desired change. Average inbreeding coefficients were about 9% at generation 5 and 19% at generation 10 for RBLUP selection in all relative desired changes, and no differences in selection between AR-BLUP and PR-BLUP were found in any generation. These results show that selection based on PR-BLUP results in similar genetic gains to that based on AR-BLUP for the three selected traits if the relative desired changes and genetic parameters are in the range of values used in the present study.